Open multimodal Llama model for strong reasoning and fast responses
Details
| Field | Value |
|---|---|
| id | meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8 |
| family | llama |
| modality | text+image->text |
| context_length | 1048576 |
| max_output_tokens | 16384 |
| release_date | 2025-04-05 |
| is_open_weights | yes |
| supports_tool_call | no |
| supports_reasoning | no |
| supports_structured_output | yes |
| supports_attachment | yes |
| input_modalities | text, image |
| output_modalities | text |